15BM60056 Sarthak Singla 15BM60085 WHAT IS VIRAL MARKETING? Viral: Circulated rapidly and widely from one internet user to another (image, video, piece of information, etc) Eg. Gangnam style, Kolaveri di etc
Viral Marketing: A method of marketing
where by consumers are encouraged to share information about companys goods or services via the internet. MATHEMATICAL MODELS Watts model provides a simple explanation for the spread of a video, but it ignores the fact that several people may send the video on to the same person. Watts Model predicts the total views of a video based on two parameters: N = initial number of people who view the video and R = the expected number of new viewers generated by a person who has just seen the video
The second model improves on Watts Model
by including the fact that some of the videos WATTS MODEL Watts assumes that at the beginning of the first period (t=1) the maker of the video "seeds" the video by getting N people to view it. Then during each time period, each new viewer is assumed to pass the video on to R new viewers. Thus there will be a total of S distinct viewers of the video where Equation 1 is true: (1) S = N + NR + NR2 +NR3+ .. If R>=1, S will be infinite. Indicating a viral video. R cannot stay greater than 1 forever, so in all likelihood R will drop after a while. Assuming that R stays constant at a value less than 1, multiply Equation 1 by R, obtaining Equation 2 : (2) RS = NR + NR2 +NR3+ .. Subtracting Equation 2 from Equation 1 yields: (3) S = N / (1-R) DRAWBACK OF WATTS MODEL Model assumes that each person reached at time t has never been reached before, which is unreasonable. For example, suppose there is a population of 1.000,000 people, and at the beginning of time t 800,000 people have seen the video. Then it seems highly unlikely that the NR t-1 new viewers the Watts Model generates at time t are all people who have not already seen the video.
If R >= 1, watts predicts an infinite number of people will see
the video, and this does not make sense A MORE COMPLEX VIRAL MARKETING MODEL Population Size = 10 million Probability that a person who sees the video will send the video on to at least one person = 0.1 Average number of people the video is sent to by one person = 20 R = 20*0.1 = 2 DEPENDENCE OF VIDEO SPREAD ON R THANK YOU